Fine Structure-Aware Sampling: A New Sampling Training Scheme for Pixel-Aligned Implicit Models in Single-View Human Reconstruction

Abstract

Pixel-aligned implicit models, such as PIFu, PIFuHD, and ICON, are used for single-view clothed human reconstruction. These models need to be trained using a sampling training scheme. Existing sampling training schemes either fail to capture thin surfaces (e.g. ears, fingers) or cause noisy artefacts in reconstructed meshes. To address these problems, we introduce Fine Structured-Aware Sampling (FSS), a new sampling training scheme to train pixel-aligned implicit models for single-view human reconstruction. FSS resolves the aforementioned problems by proactively adapting to the thickness and complexity of surfaces. In addition, unlike existing sampling training schemes, FSS shows how normals of sample points can be capitalized in the training process to improve results. Lastly, to further improve the training process, FSS proposes a mesh thickness loss signal for pixel-aligned implicit models. It becomes computationally feasible to introduce this loss once a slight reworking of the pixel-aligned implicit function framework is carried out. Our results show that our methods significantly outperform SOTA methods qualitatively and quantitatively. Our code is publicly available at https://github.com/kcyt/FSS.

Accepted in Proceedings of the AAAI Conference on Artificial Intelligence, 2024 (AAAI 2024)

References in corpus (10)

• High-Resolution Image Synthesis and Semantic Manipulation with Conditional GANs
• Deep Appearance Models for Face Rendering
• PIFu: Pixel-Aligned Implicit Function for High-Resolution Clothed Human Digitization
• PIFuHD: Multi-Level Pixel-Aligned Implicit Function for High-Resolution 3D Human Digitization
• Neural Body: Implicit Neural Representations with Structured Latent Codes for Novel View Synthesis of Dynamic Humans
• Function4D: Real-time Human Volumetric Capture from Very Sparse Consumer RGBD Sensors
• Photorealistic Monocular 3D Reconstruction of Humans Wearing Clothing
• PaMIR: Parametric Model-Conditioned Implicit Representation for Image-based Human Reconstruction
• StereoPIFu: Depth Aware Clothed Human Digitization via Stereo Vision
• IntegratedPIFu: Integrated Pixel Aligned Implicit Function for Single-view Human Reconstruction